Re: [AMBER] Protein drift from solvent box

From: David A Case <>
Date: Thu, 19 Oct 2017 12:11:11 -0400

On Thu, Oct 19, 2017, Lizelle Lubbe wrote:

I thought that 10A was commonly used for the non-bonded cutoff distance
and to avoid the computational cost of extra waters I chose 10A instead of
12A for the barrier distance.

10 Ang is pretty short; 12-14 Ang would be some better, but that's mostly a
recommendation for future work, i.e. you probably don't need to re-do the
simulations you already have.
> Could you please explain to me why this is incorrect and how it would
> cause shrinkage of the box?

What you did sounds fine.

> I measured the glycan-glycan distance between neighbouring periodic
> images after 8ns production and during the simulation it ranges between
> 25-70A. Does this mean that the protein drift is just an imaging issue
> or can the molecules start interacting later on?

The protein drift should be just an imaging problem. See if running autoimage
followed visualization looks OK.


AMBER mailing list
Received on Thu Oct 19 2017 - 09:30:04 PDT
Custom Search